The present invention relates to a navigation device which is mounted in a vehicle and uses probability in is introduced into a map matching method in order to improve navigation accuracy of the vehicle.
A navigation device is mounted in a vehicle and indicates on a display unit mounted in the vehicle which the vehicle's present location. The types of devices help the driver of the vehicle to drive it to a destination.
FIG. 1 shows a conventional navigation device 1 which is mounted in a vehicle and basically operated by a central processing unit 4 to which a direction sensor 2, a movement distance sensor 3 and so forth are connected, to obtain the presumed present location of the vehicle. A road map read from a road map information recording medium 6 such as a CD-ROM through a reader 5 (which is called a medium controller) is stored in a buffer memory 7 and then indicated on a display unit 8 together with the presumed present location. The display unit 8, operated by a display controller 9, displays road map data stored in an image memory 10. The displayed map corresponds to the present location of the vehicle and the vicinity thereof. A route
approximate to the shortest line extending from the starting place to the destination for the vehicle can also be displayed together with the road map data by manipulating the keys of a keyboard 11 connected to the central processing unit 4. As a result, the driver of the vehicle can take an optional route to the destination by viewing the destination and the shortest-line route displayed on the display unit 8.
However, the accuracy of devices such as gyroscope of a high accuracy type a terrestrial magnetism sensor and a steering angle sensor as the direction sensors vary considerably. Besides, the detection accuracy of the movement distance sensor 3 may deteriorate due to wheel slip or the like. As a result, the accuracy of the presumption of the present location of the vehicle through the navigation device 1 drops.
A map matching method in which road information is used as much as possible to judge whether or not the presumed movement loci of the vehicle are coincident with compensate for drops in sensor accuracy has received much attention. However, if there are a plurality of presumed present locations coincident with Y-crossed roads, mutually nearby right-turning roads or the like as shown in the road map in FIG. 1, it is difficult to judge which of the presumed present locations the proper one. Even if a presumed present location regarded as being of smaller presumption error is chosen as the proper one, it is often not the actual present location of the vehicle, because of accumulation in the operation of sensors. In that case, the later presumptions of the present location of the vehicle are further confused and the accuracy of the navigation device further deteriorates.